Maths to help manage flu pandemics
Tuesday, 9 October 2012
A University of Adelaide mathematician, with colleagues in the UK, has developed a model that will help public health authorities better manage future influenza pandemics and other new disease outbreaks.
The mathematical model makes use of the wide range of information recorded about infected households during a pandemic to enable improved estimates of the severity of the outbreak and the likely spread of infection.
Applied Mathematics lecturer Dr Joshua Ross said the model, which would be available to health authorities around the world, would help them produce a more informed response and aid decision-making.
"During a pandemic, government health agencies need to make decisions concerning the likely demands on the health system, whether there is a need for any social isolation measures, whether antivirals should be given and how widely," said Dr Ross.
"Our modelling provides a much clearer picture of how much infection there is within the population, with an improved understanding of the transmissibility of the disease and the probability of different clinical outcomes."
Dr Ross worked with University of Warwick mathematicians Dr Thomas House and Professor Matt Keeling and medical professionals from the UK's Health Protection Agency. They analysed early infection patterns within 424 households in Birmingham in the 2009 H1N1 ("swine flu") pandemic and their research paper, 'Estimation of outbreak severity and transmissibility: Influenza A(H1N1)pdm09 in households', is published in BMC Medicine.
The researchers used data recorded during the outbreak to estimate probabilities related to the numbers of individuals infected showing symptoms, the numbers of those with symptoms seeking medical help and having a laboratory diagnostic test, and the accuracy of the test.
"Traditionally, information on individual clinical outcomes is simplified to binary data - an individual is infected or not," Dr Ross said. "This new approach uses all these different levels of recorded data to develop a model that fits the whole process; combining this with modern statistical methods we are able to make much more accurate estimates about transmissibility and clinical outcomes.
"This would have been very valuable in the 2009 pandemic where the variable severity - often mild in nature - meant low rates of confirmation of cases. That posed problems for health authorities in determining an appropriate management response overseas and in Australia."
Dr Ross said this model would be useful in future pandemics to complement more expensive and sometimes-inconclusive blood testing as a "real-time" analysis tool to assess the true number of cases.
Dr Ross's research was funded under the Australian Research Council's Discovery Project program.